Estructura ¶

0.1 Objetivo de negocio ¶

1.1. Carga de librerias ¶


# Ignore warnings

import warnings
warnings.filterwarnings("ignore")

# General.

import pandas as pd
import numpy as np

# Visualization.

import matplotlib.pyplot as plt
import plotly.express as px
import seaborn as sns
import folium


# Train, Validation, Test split.

from sklearn.model_selection import train_test_split


# Importing the functions:

from aux_functions import *
from transformations import count_checkins

2.1. Carga de datos ¶


# Load the data business_df.parquet

business_df = pd.read_parquet('../data/processed/business_df.parquet')

# Load the data review_df.parquet

categories_df = pd.read_parquet('../data/processed/categories_df.parquet')


# Load the data checkin_df.parquet.

checkin_df = pd.read_parquet('../data/raw/yelp_academic_dataset_checkin.parquet')

# Count of checkins per business_id.

checkin_df['total_checkins'] = checkin_df['date'].apply(count_checkins)


# Verification of the shape for 'business_df'

print("business_df contiene " + str(business_df.shape[0]) +" filas y " + str(business_df.shape[1]) + " columnas")

# Verification of the shape for 'categories_df'

print("categories_df contiene " + str(categories_df.shape[0]) +" filas y " + str(categories_df.shape[1]) + " columnas")

# Verification of the shape for 'checkin_df'

print("checkin_df contiene " + str(checkin_df.shape[0]) +" filas y " + str(checkin_df.shape[1]) + " columnas")

business_df contiene 150346 filas y 99 columnas
categories_df contiene 150346 filas y 1313 columnas
checkin_df contiene 131930 filas y 3 columnas


# Merge the checkin_df with the business_df by business_id.

business_df = pd.merge(business_df, checkin_df[['business_id', 'total_checkins']], on='business_id', how='inner')

# Display the first 5 rows of the business_df

business_df.head()


# Convert the type of data for each variable from 'object' to 'float64'.

business_df['total_hours_Monday'] = business_df['total_hours_Monday'].astype('float64')
business_df['total_hours_Tuesday'] = business_df['total_hours_Tuesday'].astype('float64')
business_df['total_hours_Wednesday'] = business_df['total_hours_Wednesday'].astype('float64')
business_df['total_hours_Thursday'] = business_df['total_hours_Thursday'].astype('float64')
business_df['total_hours_Friday'] = business_df['total_hours_Friday'].astype('float64')
business_df['total_hours_Saturday'] = business_df['total_hours_Saturday'].astype('float64')
business_df['total_hours_Sunday'] = business_df['total_hours_Sunday'].astype('float64')


# Convert 'review_count' and 'total_checkins' from 'object' to 'int64'.

business_df['review_count'] = business_df['review_count'].astype('int64')
business_df['total_checkins'] = business_df['total_checkins'].astype('int64')
business_df['is_open'] = business_df['is_open'].astype('int64')


# From the 'categories_df' dataframe, convert the 'Restaurants' variable from 'object' to 'int64'.

categories_df['Restaurants'] = categories_df['Restaurants'].astype('int64')


# Extract the business_id of the Restaurants in the 'categories_df' dataframe.

ids = categories_df[categories_df['Restaurants'] == 1]['business_id'].tolist()

# Join the variables of business_df with the 'ids' list.

restaurants_df = business_df[business_df['business_id'].isin(ids)].reset_index(drop=True)

# Drop the rows with 'is_open' = 0.

restaurants_df = restaurants_df[restaurants_df['is_open'] == 1].reset_index(drop=True)

# Convert the type of data for each variable since variable 8 to from 'object' to 'int64'. If an error occurs, skip to the next variable.

for i in range(6, len(restaurants_df.columns)):
    try:
        restaurants_df.iloc[:,i] = restaurants_df.iloc[:,i].astype('float64')
    except:
        continue


# Display the first 5 rows of the restaurants_df.

restaurants_df.head()


# Verification of the different types of data in the 'restaurants_df'.

print(restaurants_df.dtypes.sort_values().to_frame('feature_type').groupby(by = 'feature_type').size().to_frame('count').reset_index())

# Print the type of data for each variable.

restaurants_df.dtypes

  feature_type  count
0      float64     87
1       object     13

business_id              object
name                     object
address                  object
city                     object
state                    object
                         ...   
total_hours_Thursday    float64
total_hours_Friday      float64
total_hours_Saturday    float64
total_hours_Sunday      float64
total_checkins          float64
Length: 100, dtype: object


uniqueValues = restaurants_df.nunique()
print('Número de valores distintos en cada variable :')
print(uniqueValues)

Número de valores distintos en cada variable :
business_id             34516
name                    23102
address                 32340
city                      839
state                      16
                        ...  
total_hours_Thursday       90
total_hours_Friday         89
total_hours_Saturday       96
total_hours_Sunday         88
total_checkins           1871
Length: 100, dtype: int64


# Graphic of the distribution of the variable 'stars'.

sns.set(style="whitegrid")
ax = sns.countplot(x="stars", data=restaurants_df, palette="Blues_d")
plt.title('Distribución de las estrellas')

plt.rcParams['figure.figsize'] = (20, 5)

plt.show()


# Replace the stars values as a binary problem:

restaurants_df['stars'] = restaurants_df['stars'].apply(lambda x: 1 if x >= 4 else 0)

# Plot the new graph.

sns.set(style="whitegrid")
ax = sns.countplot(x="stars", data=restaurants_df, palette="Blues_d")
plt.title('Distribución de las estrellas')

plt.rcParams['figure.figsize'] = (20, 5)

plt.show()

3.1. Nombramiento de datos y aspéctos éticos ¶


# Drop variables that are not relevant for the restarutants. Those ones that starts by 'attributes_HairSpecializesIn_'.

restaurants_df = restaurants_df.drop([col for col in restaurants_df.columns if col.startswith('attributes_HairSpecializesIn')], axis=1)

# Drop the variables of  'adress', 'postal_code'.

restaurants_df = restaurants_df.drop(['address', 'postal_code'], axis=1)


# Print the current shape of 'restaurants_df'.

print("El DF restaurants_df contiene " + str(restaurants_df.shape[0]) +" filas y " + str(restaurants_df.shape[1]) + " columnas")

El DF restaurants_df contiene 34516 filas y 90 columnas


# Save in 'restaurants_duplicated' the duplicated rows.

restaurants_duplicated = restaurants_df[restaurants_df.duplicated()]

# Print the number of duplicated rows in 'restaurants_duplicated' 

print('Hay un número de '+str(len(restaurants_duplicated))+ " valores duplicados en el dataset restaurants_df")

Hay un número de 0 valores duplicados en el dataset restaurants_df


# Sum the number of nulls values of the different columns and rows, then the data is sorted from highest to lowest.

pd_series_null_columns = restaurants_df.isnull().sum().sort_values(ascending=False)
pd_series_null_rows = restaurants_df.isnull().sum(axis=1).sort_values(ascending=False)

# Create two dataframes showing the null values and proportion of business_df in rows and columns.

pd_null_columnas = pd.DataFrame(pd_series_null_columns, columns=['Valores nulos en columnas'])     
pd_null_filas = pd.DataFrame(pd_series_null_rows, columns=['Valores nulos en filas'])  
pd_null_columnas['porcentaje'] = pd_null_columnas['Valores nulos en columnas']/business_df.shape[0]*100 
pd_null_filas['porcentaje_filas']= pd_null_filas['Valores nulos en filas']/business_df.shape[1]


# Show the number of null values in the rows and the proportion.

(pd_null_filas[pd_null_filas['Valores nulos en filas']>0].value_counts()).sort_values(ascending=False)

Valores nulos en filas  porcentaje_filas
35                      0.35                2493
34                      0.34                2163
36                      0.36                1993
33                      0.33                1432
37                      0.37                1420
                                            ... 
12                      0.12                  21
11                      0.11                   8
10                      0.10                   3
8                       0.08                   1
6                       0.06                   1
Length: 66, dtype: int64


# Show the variables with more than 0 null values in columns.

pd_null_columnas[pd_null_columnas['Valores nulos en columnas']>0]


# Create a DF with all the categorical variables.

categorical_variables = restaurants_df[[col for col in restaurants_df.columns if not col.startswith('total_') and not col.startswith('review_count')]]


# Create a DF with categorical_variables as rows and the number of unique values and distinct values as columns.

categorical_variables_df = pd.DataFrame(columns=['unique_values', 'distinct_values'])

for i in categorical_variables:
    categorical_variables_df.loc[i, 'unique_values'] = restaurants_df[i].nunique()
    categorical_variables_df.loc[i, 'distinct_values'] = restaurants_df[i].unique().tolist()

# Display the first 5 rows of the categorical_variables_df.

categorical_variables_df


# Group by type of operation the number of stars.

df_vegan = pd.DataFrame(restaurants_df.groupby(('attributes_DietaryRestrictions_vegan'), 
as_index=False)['stars'].sum())

# Plot an histogram of 'df_vegan'

fig = px.histogram(df_vegan, x='attributes_DietaryRestrictions_vegan', y='stars', color= 'attributes_DietaryRestrictions_vegan')
fig.show()


# Group by type of operation the number of stars.

df_vegan = pd.DataFrame(restaurants_df.groupby(('attributes_DietaryRestrictions_vegetarian'), 
as_index=False)['stars'].sum())

# Plot an histogram of 'df_vegan'

fig = px.histogram(df_vegan, x='attributes_DietaryRestrictions_vegetarian', y='stars', color= 'attributes_DietaryRestrictions_vegetarian')
fig.show()


# Group by type of operation the number of stars.

df_gluten = pd.DataFrame(restaurants_df.groupby(('attributes_DietaryRestrictions_gluten-free'), 
as_index=False)['stars'].sum())

# Plot an histogram of 'df_vegan'

fig = px.histogram(df_gluten, x='attributes_DietaryRestrictions_gluten-free', y='stars', color= 'attributes_DietaryRestrictions_gluten-free')
fig.show()


# Group by type of operation the number of stars.

df_delivery = pd.DataFrame(restaurants_df.groupby(('attributes_RestaurantsDelivery'), 
as_index=False)['stars'].sum())

# Plot an histogram of 'df_delivery'

fig = px.histogram(df_delivery, x='attributes_RestaurantsDelivery', y='stars', color= 'attributes_RestaurantsDelivery')
fig.show()


# Group by the number of stars the variables that starts by 'attributes_BusinessParking'.

business_parking = ['attributes_BusinessParkinggarage', 'attributes_BusinessParkinglot', 'attributes_BusinessParkingstreet',
                    'attributes_BusinessParkingvalidated', 'attributes_BusinessParkingvalet']


                        
df_variables = pd.DataFrame(restaurants_df.groupby((business_parking), 
as_index=False)['stars'].sum())


# Plot 2 histograms by row of 'df_diets':

for variable in business_parking:
    fig = px.histogram(df_variables, x=variable, y='stars', color= variable)
    fig.show()


# Group by type of operation the number of stars.

df_delivery = pd.DataFrame(restaurants_df.groupby(('attributes_BusinessAcceptsCreditCards'), 
as_index=False)['stars'].sum())

# Plot an histogram of 'df_delivery'

fig = px.histogram(df_delivery, x='attributes_BusinessAcceptsCreditCards', y='stars', color= 'attributes_BusinessAcceptsCreditCards')
fig.show()


# Convert latitude and longitude to float64.

restaurants_df['latitude'] = restaurants_df['latitude'].astype('float64')
restaurants_df['longitude'] = restaurants_df['longitude'].astype('float64')

# Select the geographical coordinates and the name of the restaurants.

restaurants = restaurants_df[['business_id','latitude','longitude','name','stars', 'state']]

# Set the mapbox access token.

mapbox_access_token = 'pk.eyJ1IjoiYWxtYXNzbml5YW1hdCIsImEiOiJjamQ3NGY2Zms0emhmMnFuMjJ5OGNvOWoxIn0.DKGZcDRHVFRYNiQBe1D_zw'

# Create scatterplot in the map.

fig = px.scatter_mapbox(restaurants,
                        lat='latitude',
                        lon='longitude',
                        hover_name='name',
                        hover_data=['stars', 'state'],
                        zoom=3,
                        height=900,
                        width=900,
                        color='stars',
                        color_continuous_scale=px.colors.sequential.Bluered,
                        labels={'stars':'Stars'},
                        )

fig.update_coloraxes(colorbar=dict(title='Stars', tickvals=[0,1], ticktext=['0','1']))

fig.update_layout(mapbox_style="open-street-map",
                 margin={"r":0,"t":0,"l":0,"b":0}
                 )


# Display the map.

fig.show()


# Drop the rows with the state 'AB'.

restaurants_df = restaurants_df[restaurants_df['state'] != 'AB']


# Group by the state of the restaurants and show in one column the total of stars and in other column the total of restaurants.

df_stars = pd.DataFrame(restaurants_df.groupby(('state'),
as_index=False)['stars'].sum()).sort_values(by='stars', ascending=False).reset_index(drop=True)

df_stars['total_restaurants'] = restaurants_df.groupby(('state'),
as_index=False)['stars'].count()['stars']



# Display the first 5 rows of the df_stars.

df_stars.head(3)


# Import the json file.

geo_data = '../data/maps/us-states.json' 

# Create a map object.

map = folium.Map(location=[37.0902, -95.7129], zoom_start=4)

# Create a choropleth map.

folium.Choropleth(
    geo_data=geo_data,
    name='choropleth',
    data=df_stars,
    columns=['state', 'stars'],
    key_on='feature.id',
    fill_color='YlGn',
    fill_opacity=0.7,
    line_opacity=0.2,
    legend_name='Total of stars'
).add_to(map)

# Plot the map.

map


# Create a list with the rows that will be included in the new dataframe.

rows = []

# Loop through all pairs of variables.

for var1 in categorical_variables:
  col = []
  for var2 in categorical_variables:

    # Calculate Cramer's V for the pair of variables.

    observed = pd.crosstab(categorical_variables[var1], categorical_variables[var2])
    if observed.size == 0:
      col.append(None)
      continue
    cramers = cramers_V(categorical_variables[var1], categorical_variables[var2])
    col.append(round(cramers,2))
    
  rows.append(col)


# Creation of the Cramer's V matrix.

cramers_results = np.array(rows)
matrix = pd.DataFrame(cramers_results, columns = categorical_variables.columns, index =categorical_variables.columns)

# Convert as index the 0 column.

matrix = matrix.set_index(matrix.columns[0])

# Display the Cramer's V matrix with all columns and rows.

matrix


# Drop the 'business_id', 'name' and 'city' columns.

matrix = matrix.drop(columns=['business_id', 'name', 'city','is_open'])

# Drop the 'business_id', 'name' and 'city' rows.

matrix = matrix.drop(index=['business_id', 'name', 'city','is_open'])


# Display the Cramer's V matrix with all columns and rows.

matrix


# Export to .csv.

matrix.to_csv("../data/processed/matrix_crammer.csv")


# Select the numeric variables to save them in the new dataframe 'num_var_df'. Variables that start by 'total_' and 'review_count'.

num_var_df = restaurants_df[[col for col in restaurants_df.columns if col.startswith('total_') or col.startswith('review_count')]]


# Display the correlation matrix of the 'num_var_df'. 

get_corr_matrix(dataset = num_var_df, metodo='pearson', size_figure=[20,15])


# Obtain the outliers of the 'num_var_df' dataset.

outliers = get_outliers(num_var_df)

# Display number of rows withous NaN values.

print(outliers.count())

# Boxplot of the 'num_var_df' dataset.

fig = px.box(num_var_df, y=num_var_df.columns)
fig.show()

review_count             3368
total_hours_Monday          0
total_hours_Tuesday      1795
total_hours_Wednesday    1799
total_hours_Thursday     1836
total_hours_Friday       1682
total_hours_Saturday     1674
total_hours_Sunday       1783
total_checkins           3850
dtype: int64


# Export 'restaurants_df' as csv in the folder of processed data.

restaurants_df.to_parquet("../data/processed/restaurants_df.parquet")

Ir arriba del documento

	business_id	name	address	city	state	postal_code	latitude	longitude	stars	review_count	...	attributes_DietaryRestrictions_soy-free	attributes_DietaryRestrictions_vegetarian	total_hours_Monday	total_hours_Tuesday	total_hours_Wednesday	total_hours_Thursday	total_hours_Friday	total_hours_Saturday	total_hours_Sunday	total_checkins
0	Pns2l4eNsfO8kk83dixA6A	Abby Rappoport, LAC, CMQ	1616 Chapala St, Ste 2	Santa Barbara	CA	93101	34.4266787	-119.7111968	5.0	7	...	nan	nan	nan	nan	nan	nan	nan	nan	nan	1
1	mpf3x-BjTdTEA3yCZrAYPw	The UPS Store	87 Grasso Plaza Shopping Center	Affton	MO	63123	38.551126	-90.335695	3.0	15	...	nan	nan	0.0	10.5	10.5	10.5	10.5	6.0	nan	15
2	tUFrWirKiKi_TAnsVWINQQ	Target	5255 E Broadway Blvd	Tucson	AZ	85711	32.223236	-110.880452	3.5	22	...	nan	nan	14.0	14.0	14.0	14.0	15.0	15.0	14.0	401
3	MTSW4McQd7CbVtyjqoe9mw	St Honore Pastries	935 Race St	Philadelphia	PA	19107	39.9555052	-75.1555641	4.0	80	...	nan	nan	13.0	13.0	13.0	13.0	14.0	14.0	14.0	335
4	mWMc6_wTdE0EUBKIGXDVfA	Perkiomen Valley Brewery	101 Walnut St	Green Lane	PA	18054	40.3381827	-75.4716585	4.5	13	...	nan	nan	nan	nan	8.0	6.0	10.0	10.0	6.0	14

	business_id	name	address	city	state	postal_code	latitude	longitude	stars	review_count	...	attributes_DietaryRestrictions_soy-free	attributes_DietaryRestrictions_vegetarian	total_hours_Monday	total_hours_Tuesday	total_hours_Wednesday	total_hours_Thursday	total_hours_Friday	total_hours_Saturday	total_hours_Sunday	total_checkins
0	MTSW4McQd7CbVtyjqoe9mw	St Honore Pastries	935 Race St	Philadelphia	PA	19107	39.955505	-75.155564	4.0	80.0	...	NaN	NaN	13.0	13.0	13.0	13.0	14.0	14.0	14.0	335.0
1	CF33F8-E6oudUQ46HnavjQ	Sonic Drive-In	615 S Main St	Ashland City	TN	37015	36.269593	-87.058943	2.0	6.0	...	NaN	NaN	0.0	16.0	16.0	16.0	15.0	13.0	14.0	22.0
2	bBDDEgkFA1Otx9Lfe7BZUQ	Sonic Drive-In	2312 Dickerson Pike	Nashville	TN	37207	36.208102	-86.768170	1.5	10.0	...	NaN	NaN	0.0	15.0	15.0	10.0	10.0	11.0	15.0	21.0
3	eEOYSgkmpB90uNA7lDOMRA	Vietnamese Food Truck		Tampa Bay	FL	33602	27.955269	-82.456320	4.0	10.0	...	NaN	NaN	3.0	3.0	3.0	3.0	3.0	5.0	3.0	4.0
4	il_Ro8jwPlHresjw9EGmBg	Denny's	8901 US 31 S	Indianapolis	IN	46227	39.637133	-86.127217	2.5	28.0	...	NaN	NaN	16.0	16.0	16.0	16.0	16.0	16.0	16.0	54.0

	Valores nulos en columnas	porcentaje
attributes_AcceptsInsurance	34501	26.150989
attributes_RestaurantsCounterService	34500	26.150231
attributes_DietaryRestrictions_dairy-free	34498	26.148715
attributes_DietaryRestrictions_kosher	34498	26.148715
attributes_DietaryRestrictions_vegan	34498	26.148715
...	...	...
total_hours_Saturday	4041	3.062988
total_hours_Wednesday	4023	3.049344
total_hours_Thursday	3550	2.690821
total_hours_Friday	3385	2.565755
attributes_RestaurantsTakeOut	2854	2.163268

	state	stars	total_restaurants
0	PA	3730	1775
1	FL	2945	660
2	TN	1385	692

	business_id	name	city	state	latitude	longitude	stars	is_open	attributes_ByAppointmentOnly	attributes_BusinessAcceptsCreditCards	...	attributes_BestNights_thursday	attributes_BestNights_sunday	attributes_BestNights_saturday	attributes_DietaryRestrictions_dairy-free	attributes_DietaryRestrictions_gluten-free	attributes_DietaryRestrictions_vegan	attributes_DietaryRestrictions_kosher	attributes_DietaryRestrictions_halal	attributes_DietaryRestrictions_soy-free	attributes_DietaryRestrictions_vegetarian
business_id	1.0	1.0	1.0	1.0	1.0	1.0	1.0	NaN	1.0	1.0	...	1.0	1.0	1.0	1.0	1.0	1.0	NaN	1.0	1.0	1.0
name	1.0	1.0	0.68	0.78	0.97	0.95	0.88	NaN	0.98	0.94	...	0.87	0.82	0.88	1.0	1.0	1.0	NaN	1.0	1.0	1.0
city	1.0	0.68	1.0	1.0	0.99	0.98	0.05	NaN	0.11	0.07	...	0.12	0.1	0.12	1.0	1.0	1.0	NaN	1.0	1.0	1.0
state	1.0	0.78	1.0	1.0	1.0	1.0	0.01	NaN	0.01	0.02	...	0.02	0.01	0.01	0.41	1.0	0.57	NaN	1.0	0.47	0.45
latitude	1.0	0.97	0.99	1.0	1.0	1.0	0.98	NaN	0.99	0.98	...	1.0	0.99	1.0	1.0	1.0	1.0	NaN	1.0	1.0	1.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
attributes_DietaryRestrictions_vegan	1.0	1.0	1.0	0.57	1.0	1.0	0.0	NaN	NaN	0.0	...	NaN	NaN	NaN	0.28	0.17	0.78	NaN	0.0	0.17	0.42
attributes_DietaryRestrictions_kosher	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
attributes_DietaryRestrictions_halal	1.0	1.0	1.0	1.0	1.0	1.0	0.0	NaN	NaN	0.0	...	NaN	NaN	NaN	0.0	0.05	0.0	NaN	0.22	0.0	0.0
attributes_DietaryRestrictions_soy-free	1.0	1.0	1.0	0.47	1.0	1.0	0.09	NaN	NaN	0.0	...	NaN	NaN	NaN	0.43	0.0	0.17	NaN	0.0	0.64	0.17
attributes_DietaryRestrictions_vegetarian	1.0	1.0	1.0	0.45	1.0	1.0	0.0	NaN	NaN	0.0	...	NaN	NaN	NaN	0.28	0.01	0.42	NaN	0.0	0.17	0.78

Estructura ¶

0.1 Objetivo de negocio ¶

1.1. Carga de librerias ¶

2.1. Carga de datos ¶

3.1. Nombramiento de datos y aspéctos éticos ¶

01. EDA

0. Introducción

1. Librerias utilizadas y entorno.

1.2. Carga de funciones¶

2. Tratamiento de datos

2.2. Unión de dataframes¶

2.3. Análisis de tipos de datos¶

2.4. Diccionario de datos¶

2.5. Variable Objetivo¶

3. Análisis EDA

3.2 Análisis de valores duplicados¶

3.3. Análisis de valores nulos¶

3.3.1 Análisis de valores nulos por filas y columnas¶

3.4. Análisis de variables categóricas y numéricas¶

3.4.1 Análisis de los valores de las Variables categóricas¶

Análisis de comida para veganos. ¿Afectaría al número de estrellas tener o no este tipo de comida?¶

Análisis de comida para vegatarianos. ¿Afectaría al número de estrellas tener o no comida vegana?¶

Análisis de Celiacos. Si acude algún comensal que sea celiaco y el restaurante no tiene la capacidad de darle un correcto servicio, ¿afectaría a su valoración?¶

Análisis de servicio de reparto. ¿El servicio de reparto afectaría al número de estrellas del restaurante?¶

Análisis de Parking. Si los clientes cuentan con dificultades a la hora de apartar, ¿afectaría a la valoración?¶

Análisis de medio de pago. Si el cliente no puede pagar con tarjeta, ¿Afectaría significativamente?¶

Análisis por zona geográfica. ¿En qué estados se encuentran los restaurantes registrados?¶

Perfil más característico de restaurantes que contarían mayor número de estrellas.¶

Relación entre variables categóricas con V de Crammer y correlación¶

3.4.2 Análisis de las variables numéricas¶

Correlación entre variables numéricas¶

3.5. Análisis de Outliers¶

4. Exportación del dataset transformado

	unique_values	distinct_values
business_id	34516	[MTSW4McQd7CbVtyjqoe9mw, CF33F8-E6oudUQ46Hnavj...
name	23102	[St Honore Pastries, Sonic Drive-In, Vietnames...
city	839	[Philadelphia, Ashland City, Nashville, Tampa ...
state	16	[PA, TN, FL, IN, NV, IL, NJ, CA, LA, DE, MO, A...
latitude	33564	[39.9555052, 36.269593, 36.2081024, 27.9552692...
...	...	...
attributes_DietaryRestrictions_vegan	2	[nan, 1.0, 0.0]
attributes_DietaryRestrictions_kosher	1	[nan, 0.0]
attributes_DietaryRestrictions_halal	2	[nan, 0.0, 1.0]
attributes_DietaryRestrictions_soy-free	2	[nan, 1.0, 0.0]
attributes_DietaryRestrictions_vegetarian	2	[nan, 1.0, 0.0]